FOAF

[Jure Sah]

Hello,

I am trying to get in touch with anyone who used to work on this or is
perhaps still working on it. All the records I find online are rather
old and there are no people to contact anywhere.

Anyhow, I am wondering why FOAF never became more popular and am
considering various forms of it's practical usage. I have many questions
about the specification that I never see properly answered on the web
pages online, thus I will ask some of them here.

First of all, a single FOAF file is intended to describe a single
person, right? I can see many situations, where more than one FOAF file
could be present online to describe a single person (for example when
FOAF data is automatically extracted from other data sources). How in
that case, would it be appropriate to link the different FOAF files
together?

Second, I am thinking of the case where a FOAF file could be used to
publish public information about a person from an official source (for
example a list of FOAF files defining the students of a college). In
this case, some kind of certificate would have to be used to confirm the
institution providing the data is to be held responsible for the data's
accuracy and that the data has been untampered with after the
institution has published it. How does one combine FOAF and certificates
in a correct manner?

Third, in the examples of FOAF files I have seen online, I notice the
primary structure of the FOAF web is the <foaf:knows>. How would one
correctly use FOAF to describe different kinds of relationships (for
example: friends, colleagues, schoolmates) without confusing any
existing FOAF browsing software?

Thank you for any answers. :)

--
Primary function: Coprocessor
Secondary function: Cluster commander

http://www.thought-beacon.net

Pay once per lifetime webhosting:
http://farcomm-it.com/?ref=jsah

We are the paragon of humanity. You may worship us. From afar.

01010010 01100101 01110011 01101001 01100100 01100101 01101110 01110100
01000010 01000001 01010011 01001001 01000011

[Andy Dingley]

Jure Sah wrote:

I am trying to get in touch with anyone who used to work on this or is
perhaps still working on it. All the records I find online are rather
old and there are no people to contact anywhere.

All the people are still there, and generally still in Bristol, if not
at ILRT (try buying beer in the Highbury Vaults!)

Anyhow, I am wondering why FOAF never became more popular

It did, but like most SemWeb stuff it's hidden "under the hood".

First of all, a single FOAF file is intended to describe a single
person, right?

FOAF is RDF, so frees itself from XML concepts like "file". It's a
small problem to manipulate multiple entities in one document and a big
problem to manage one entity described across many documents. RDF tools
solve nearly all of this for you, and transparently. This is why it's
important to use something like Jena rather than document-based XML
tools.

The rest is just an exercise in web searching. Try identifying the
obvious author names, then tracking down their current projects.

[Jure Sah]

Andy Dingley pravi:

>I am trying to get in touch with anyone who used to work on this or is
perhaps still working on it. All the records I find online are rather
old and there are no people to contact anywhere.

All the people are still there, and generally still in Bristol, if not
at ILRT (try buying beer in the Highbury Vaults!)

Which country already?

>Anyhow, I am wondering why FOAF never became more popular

It did, but like most SemWeb stuff it's hidden "under the hood".

....and unused.

>First of all, a single FOAF file is intended to describe a single
person, right?

FOAF is RDF, so frees itself from XML concepts like "file". It's a
small problem to manipulate multiple entities in one document and a big
problem to manage one entity described across many documents. RDF tools
solve nearly all of this for you, and transparently. This is why it's
important to use something like Jena rather than document-based XML
tools.

You're asking me to use a Java library? Hmm, right, I think I understand
why all my friends hate XML now...

Not to mention, it really doesn't answer my question. I don't use "XML
tools", document or otherwise, unless you count VI as one. I simply
recognize the possibility of there being more than one source of
information for describing a certain someone. The obvious solution is to
use a FAT (or should I say XAT? *g*), but how exactly does one use RDF
to form this kind of structure?

The rest is just an exercise in web searching. Try identifying the
obvious author names, then tracking down their current projects.

You could have just summed up your entire post into "RTFM" then.

--
Primary function: Coprocessor
Secondary function: Cluster commander

http://www.thought-beacon.net

Pay once per lifetime webhosting:
http://farcomm-it.com/?ref=jsah

We are the paragon of humanity. You may worship us. From afar.

01010010 01100101 01110011 01101001 01100100 01100101 01101110 01110100
01000010 01000001 01010011 01001001 01000011

[Andy Dingley]

Jure Sah wrote:

You're asking me to use a Java library? Hmm, right, I think I understand
why all my friends hate XML now...

XML has processing tools in almost any language. RDF is a couple of
years newer and has been developed by fewer people, so there are as yet
fewer language bindings for it. If you want one, write one.

The popular stuff is in Java, This is because:

!. All the cool kids were using Java

2. Jena is in Java.

3. Protege is in Java.

As Java works, is popular, and the two favourite RDF tools are using
it, then why move away ? If you use Redland instead of Jena then
you've access from C++ or Python. You're welcome to use anything else
you fancy, it's just that the rest of us are already happily sat in one
corner.

Not to mention, it really doesn't answer my question. I don't use "XML
tools", document or otherwise, unless you count VI as one.

If you insist on only seeing the world as a linear text document
through a fairly crude character-based editor, then you're never going
to get very far managing triples. Things are harder now, user
interfaces are fatter. If we're going to work with this world, then
we're going to need fatter tools. Download a copy of Protege and have a
play with some of the newer plugins for that.

You could have just summed up your entire post into "RTFM" then.

Really I wish I had.

[Jure Sah]

Andy Dingley pravi:

As Java works, is popular, and the two favourite RDF tools are using
it, then why move away ? If you use Redland instead of Jena then
you've access from C++ or Python. You're welcome to use anything else
you fancy, it's just that the rest of us are already happily sat in one
corner.

Because nowadays everybody is using some high level language, which is
based on another high level language, which is based on another high
level language, which is based on another high level language, which is
based on another high level language, which is based on another high
level language, which is based on another high level language, which is
based on another high level language. And the code at the core is bloated.

Or that is the definition of anything running on a computer nowadays.
I've been working in the area of artificial intelligence for 3 years and
I have come to realize the only reason why many things "can't be done"
is not because the computers aren't up to it, it's because all the code
we have today are workarounds around workarounds around workarounds that
don't really need to be there in the first place.

In short: You're missing the world of what can be done and limiting the
capabilities of future implementations by getting at this problem the
way you are.

>Not to mention, it really doesn't answer my question. I don't use "XML
tools", document or otherwise, unless you count VI as one.

If you insist on only seeing the world as a linear text document
through a fairly crude character-based editor, then you're never going
to get very far managing triples. Things are harder now, user
interfaces are fatter. If we're going to work with this world, then
we're going to need fatter tools. Download a copy of Protege and have a
play with some of the newer plugins for that.

It's not one bit about the editing tool, mind you. If you really knew
what you were doing, you could do FOAF, RDF or XML with a pencil onto a
piece of paper just fine. But then again you don't find people who
really know their stuff enough to be capable of explaining it, yellin'
"RTFM" around whenever asked for assistance.

Taking your abstraction levels in there you're just practicing narrow
vision yourself.

--
Primary function: Coprocessor
Secondary function: Cluster commander

http://www.thought-beacon.net

Pay once per lifetime webhosting:
http://farcomm-it.com/?ref=jsah

We are the paragon of humanity. You may worship us. From afar.

01010010 01100101 01110011 01101001 01100100 01100101 01101110 01110100
01000010 01000001 01010011 01001001 01000011

[Andy Dingley]

On Wed, 02 Aug 2006 20:06:18 +0200, Jure Sah <ad***@thought-beacon.net>
wrote:

>I've been working in the area of artificial intelligence for 3 years and
I have come to realize the only reason why many things "can't be done"
is not because the computers aren't up to it, it's because all the code
we have today are workarounds around workarounds around workarounds that
don't really need to be there in the first place.

So write everything from scratch in assembler. Or save a little effort,
pick a nice bytecode language to work in (an interpreted machinecode is
still prety efficient) and write it in Java bytecode.

If it's just _efficiency_ that's the problem, then wait a few years and
let Moore's Law catch up.

Or maybe you've got entirely the wrong end of the stick, and you're just
wondering why everyone else is at the other end of it ?

--
Cats have nine lives, which is why they rarely post to Usenet.

[pascal.grouselle]

Jure Sah a écrit :

Third, in the examples of FOAF files I have seen online, I notice the
primary structure of the FOAF web is the <foaf:knows>. How would one
correctly use FOAF to describe different kinds of relationships (for
example: friends, colleagues, schoolmates) without confusing any
existing FOAF browsing software?

You need to use another vocabulary : http://vocab.org/relationship/
within your FOAF file. With this vocabulary, you can say : this person
whom I know is a colleague, or a neighbor of mine, etc.

To see how to declare this vocabulary in the beginning of your FOAF and
how to use it, have a look on my FOAF file ;-)
http://perso.orange.fr/pascal.grouselle/foaf.rdf